Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

st.dataframe does not handle StringDtype in Pandas >= 1.0.0

See original GitHub issue

Summary

Displaying a DataFrame with a ‘string’ or StringArray dtype column fails with a NotImplementedError. This dtype is new in Pandas 1.0.0.

Steps to reproduce

Run this app:

import pandas as pd
import streamlit as st

st.header('DataFrame with StringDtype')
df = pd.DataFrame({
    'a': pd.Series('we are the knights who say ni'.split(), dtype='string'),
})
st.dataframe(df)

Debug info

Streamlit version: 0.54.0
Python version: 3.7.6
Pandas version: 1.0.0
Using Pipenv
OS version: MacOS 10.15.3

Issue Analytics

State:
Created 4 years ago
Comments:13 (3 by maintainers)

Top GitHub Comments

2reactions

JohnReidcommented, Jul 23, 2020

@jrhone labelling this as an “enhancement” might not be strong enough. These new dtypes are the preferred dtypes for people using pandas 1. Not to support them out of the box looks like a bug to me.

1reaction

nthmostcommented, Jul 25, 2020

@JohnReid Point taken.

We use labels to manage our work pipelines, but every once in a while an issue like this one sort of defines clean classification. Regardless, I think there’s a strong case for prioritizing this one. Thanks for the bump!